This invention pertains to a method and composition for purging leukemia cells from hematopoietic stem cells.
For many leukemia patients, the only hope for cure or long term survival is a bone marrow transplant from a donor. The cells for these xe2x80x9callogeneicxe2x80x9d transplants are generally obtained from a sibling or from an unrelated, HLA-matched donor. However only a small fraction of potential patients receive such transplants due to constraints such as advanced age or the lack of a matching donor. Therefore, xe2x80x9cautologousxe2x80x9d marrow transplants are also used at times, the re-infusion of cells from the patient""s own bone marrow following chemotherapy or radiation therapy that otherwise destroys the patient""s marrow. Bone marrow is taken from a patient prior to high dose chemotherapy or radiation therapy, and is later reinfused to xe2x80x9crescuexe2x80x9d the patient following the otherwise lethal therapy. However, there is currently no effective procedure to completely remove contaminating leukemia cells from bone marrow or other stem cell preparations. On the one hand, an autologous transplant does not carry the risk of short- or long-term graft-versus-host-disease, since the patient receives back his or her own bone marrow. However, a major drawback of existing autologous techniques is the lack of an effective way to remove all contaminating cancer cells from the bone marrow ex vivo. Relapses frequently result from such contaminating cancer cells. In principle, autologous transplants should be superior to allogeneic transplants if a method could be found to completely purge the transplanted cells of contaminating leukemia cells, because the risk of host-versus-graft disease would be eliminated.
Progress has been made in reducing relapse rates in autologous transplants. Various purging procedures have been used to selectively remove leukemia cells from bone marrow, such as ex vivo chemotherapy with 4-hydrocyclophosphamide, or fractionation of cells by size. The combination of these two techniques, ex vivo chemotherapy with fractionation of cells by size, would probably be considered the current state-of-the art purging procedure by most researchers. However, this combination therapy often does not completely purge leukemic cells from the transplanted material.
So-called hematopoietic xe2x80x9cstem cellsxe2x80x9d are critical in reconstituting a destroyed immune system and reconstituting red blood cell synthesis. Stem cells express the CD34 molecular marker on the cell surface. A different approach to purging bone marrow of leukemia cells is to select for cells expressing the CD34 molecule, for example by affinity chromatography using monoclonal anti-CD34 antibodies. This selection reduces the number of malignant cells, although it does not eliminate them entirely. A patient infused with the CD34 fraction of bone marrow cells will reconstitute hematopoiesis completely. CD34 selection of hematopoietic stem cells is currently being explored as an alternative purging system in autologous transplants.
Although a several log-fold reduction in leukemia cell number has been seen with these prior purging procedures, significant numbers of malignant cells are still present after purging. Currently there is no purging procedure that can completely eradicate leukemic cells from contaminated bone marrow. There is a continuing need for improved methods to efficiently purge bone marrow cells of leukemia cells.
P. Seth et al., xe2x80x9cA recombinant adenovirus expressing wild type p53 induces apoptosis in drug-resistant human breast cancer cells: A gene therapy approach for drug-resistant cancers,xe2x80x9d Cancer Gene Ther., vol. 4, pp. 383-390 (1997) discloses that a recombinant adenovirus expressing the wild type tumor suppressor gene p53 induced apoptosis in two drug-resistant human breast cancer cell lines, and that the recombinant virus selectively induced apoptosis in the breast cancer cells when the latter were mixed with CD34 cells. The differential targeting resulted from the fact that breast cancer cells express high levels of adenovirus receptors, while bone marrow cells are deficient in adenovirus receptors. See also P. Seth et al., xe2x80x9cAdenovirus-mediated gene transfer to human breast tumor cells: An approach for cancer gene therapy and bone marrow purging,xe2x80x9d J. Cancer Res., vol. 56, pp. 1346-1351 (1996).
Similarly, L. Chen et al., xe2x80x9cSelective transgene expression for detection and elimination of contaminating carcinoma cells in hematopoietic stem cell sources,xe2x80x9d J. Clin. Invest., vol. 98, pp. 2539-2548 (1996) discloses the use of replication-defective adenoviral vectors to selectively transduce breast cancer cells in the presence of bone marrow cells. Using an adenoviral vector to transduce the HSV-tk gene allowed the selective killing of breast cancer cells with ganciclovir with little effect on CFU-GM and BFU-E formulation or on long term culture initiating cells. See also J. Wilson, xe2x80x9cWhen bad gene transfer is good,xe2x80x9d J. Clin. Invest., vol. 98, p. 2435 (1996).
Unfortunately, adenovirus-based vectors are not well-suited for treating leukemias because adenovirus uptake by leukemia cells is poor.
We have discovered a gene therapy system that selectively kills leukemia cells in bone marrow, while leaving stem cells unaffected. All cells in a mixture of stem cells and leukemia cells are transfected with a high efficiency gene transfer vector. The vector carries a eukaryotic expression construct encoding a toxin gene. This toxin gene is expressed only in leukemia cells, not in stem cells. Differential expression of the toxin gene in leukemia cells and stem cells may be achieved by placing the coding sequence under the control of an appropriate promoter, such as the RSV promoter or the SV40 promoter. To the inventors"" knowledge, there have been no prior reports of specific differential regulation of genes in leukemia cells and stem cells.
Using a molecular conjugate vector, we have demonstrated high gene expression in a panel of transformed leukemia cell lines, but no gene expression in transformed, CD34-selected, primary human stem cells. We have demonstrated high levels of gene expression in leukemia cells, and essentially no expression in stem cells. We have also demonstrated the selective killing of leukemia cells mixed with a population of CD34 bone marrow cells.
The novel treatment will be useful not only for leukemia patients, but also for other cancer patients undergoing autologous bone marrow transplants (e.g., breast or lymphoma cancer patients).
The coding sequence may either directly encode a toxin, or it may encode a conditionally toxic peptide such as HSV-tk. The HSV-tk gene product is conditionally lethal, as that product is an enzyme that converts ganciclovir into a toxin. (Other toxins that could be used in this invention are discussed below.)
In a clinical setting,, a construct in accordance with this invention is transferred with high efficiency into a suspension of a human patient""s bone marrow cells, containing both the patient""s stem cells and leukemia cells. Since expression of the toxin gene is restricted to leukemic cells, only those cells will be eliminated. Since the toxin gene is not expressed in hematopoietic stem cells, the stem cells survive the purging procedure. The novel purging system can be used in autologous bone marrow transplants either by itself, or in conjunction with other purging systems. The combination of this system with others (e.g. ex vivo chemotherapy, fractionation of cells by size, selection for the CD34 marker) should be able to eliminate leukemic cells from stem cells completely. The construct is preferably designed so that the trans-gene is present only transiently, and does not integrate into the genome of the stem cells.
Experimental Results.